Mammography is a medical imaging technique used to screen for breast cancer and other abnormalities in breast tissue. Traditionally, mammography images, referred to as mammograms, have been 2D images. Unlike the conventional mammogram, X-ray tomosynthesis is a mammography technique that creates a 3D representation of the breast. See, e.g., US2006/0098855. A typical mode for viewing such a 3D model is by browsing a series of (parallel) 2D images, referred to as “a stack.” A 2D image in a stack is also referred to as a slice.
The diagnostic viewing of digital mammography images is typically performed in a Picture Archiving and Communication System (PACS). A main task of the PACS is to provide a highly efficient workflow for the health professional to review the images. The desire for high diagnostic throughput is particularly pertinent in the case of mammography screening, where an entire population of women can undergo mammography imaging. Within a few seconds, the reading physician should be able to come to the conclusion whether the breast is healthy or contains suspicious findings that should be further analyzed. When tomosynthesis images are used, a whole stack of 2D images must be reviewed for each breast instead, in contrast to conventional review of just a single 2D image. The tomosynthesis technique may potentially multiply the required review time by the number of images in the stack. Therefore, it is of great interest to create navigation tools for tomosynthesis images that can increase the speed of the review process.
One conventionally important part of the diagnostic assessment of the mammography images is to compare two breast images. To do so, the right and left breast are typically displayed simultaneously in a mirrored setup to allow a relatively simple visual comparison. Typically, the radiologist looks for symmetry between the right and left breast. Asymmetry could suggest that the images should be reviewed closer. A comparison of the new image(s) with older images is also highly useful and is common practice if older images exist.
The ability to have automatic geometric position synchronization between different stacks is a known PACS function, in a Sectra® PACS product/system it is known as the “Localizer”. It is believed that this feature is limited to stacks that have a known geometric relation, typically several scans in between which the patient did not move, which in the DICOM standard is denoted by the “Frame of reference” attribute. Geometric position synchronization of stacks without limitations is also a conventional Sectra® PACS function, but the user manually defines a common reference point for the two data sets.
Despite the above, there is a need to provide systems and methods that can make comparisons between reference images and tomosynthesis stacks to improve the accuracy, speed and/or quality of the diagnostic assessment.